Verification and implementation

When it comes to formal arms control, ‘verification’ is one of the most prominent concepts. In general terms, it refers to the processes and mechanisms used to ensure that parties (usually states) comply with the terms of the respective arms control agreement. This usually includes monitoring (e.g. with satellites or seismic monitoring devices), inspections and visits by foreign inspectors (sometimes planned well in advance, sometimes spontaneous), data sharing with inspectors (which they can then compare with their own findings) and other measures to detect and prevent violations.

In an ideal process, verification promotes trust and transparency between the signatories. It aims to provide confidence that all parties are fulfilling their commitments, that all parties are compliant and that no party is secretly violating the agreement. The aim of verification is therefore to help make arms control possible in the first place. It does so by protecting the states from having their willingness to cooperate exploited. By verifying whether the other side is adhering to the agreement they entered into, each side can engage in cooperation with peace of mind. However, even a very intrusive verification regime cannot guarantee compliant behaviour of the other side with 100-percent certainty. That being said, at the very least, extensive verification measures can drive up the costs of secretly violating the agreement. But how much verification is feasible and acceptable? Visit the other learning units to see how the issue of verification is handled in the respective field.

In general, verification walks a fine line between politically desirable transparency and politically undesirable espionage. Therefore, a minimum of trust is also necessary for verification.¹

This means that even the most sophisticated and stringent verification measures will not create the trust necessary for arms control if the starting point is complete mistrust and the assumption that states would only enter into agreements in order to exploit them unilaterally. The findings of research on verification theory are therefore paradoxical. Intrusive, far-reaching verification measures can apparently be used when a certain relationship of trust already exists, i.e. when intrusive and far-reaching verification measures are actually no longer necessary. Conversely, such measures would in fact be important where they are least likely because there is insufficient trust to introduce them.

Even if a base level of trust does exist, a number of other challenges make verification difficult. Some activities are difficult to detect, especially if they are deliberately hidden and the existing verification mechanisms intentionally undermined. Others are difficult to interpret, especially if dual-use technology is used and agreements are vague as to what activities are permitted or prohibited. Lastly, verification is expensive and requires the latest technologies to keep pace with innovation in weapons technology.

States have developed a range of methods to overcome mistrust and enable verification. These include the use of national technical means (NTMs) for independent verification, e.g. through satellites or airborne sensors; an agreement to share reports and data; and on-site inspections to verify compliance on the ground. In some cases, international organisations were founded to ensure third-party verification, as in the case of the OPCW, IAEA and CTBTO. A relatively new idea is to engage civil society and use the invisible college of scholars and activists for societal verification. However, the restrictions civil society faces in some countries significantly limit their utility. Lastly, in some cases, states have decided to dispense with verification altogether, e.g. in the area of biological weapons, because the effort would be entirely disproportionate to the result. It would simply not be possible to control all state and private facilities where biological weapons could be produced.